Detecting disease through breathing while sleeping...Nature sub-issue: Artificial intelligence provides digital biomarkers for Parkinson's disease diagnosis

Parkinson's disease (PD) is one of the fastest growing diseases in the world and the second most common neurological disease after Alzheimer's disease

Over the years, researchers in the field have discovered several potential Parkinson's disease biomarkers, such as cerebrospinal fluid, blood biochemistry, and neuroimaging that have good potential for detection

Additionally, to allow the system to operate every night at home while sleeping, without touching the subject's body, the team developed a device that looks like a home Wi-Fi router, but doesn't provide internet access , instead, emit radio signals and analyze their reflections on the surrounding environment to extract the breathing characteristics of the subject

An important part of the model's design is that it learns from nighttime breathing the auxiliary task of predicting a person's quantitative electroencephalogram (qEEG), which prevents the model from overfitting and helps explain the model's output

In conclusion, this study demonstrates that advances in artificial intelligence can lead to the development of new biomarkers to support medical development by addressing important unsolved challenges in neuroscience research